Bestia
Real estate lead and property CRM (Bestia / Leadboard Lite): leads list with filters and search, lead & property detail with tabbed data (features, ownership, tax, mortgage, price history), property comparison, notes, contacts, and multi-channel communication. Showcased at LA Tech Week — 300 agents on the waiting list.
My Role
UI/UX Designer & Frontend Developer
Core Stack
Next.js 15, React 19, TypeScript, Figma, Tailwind 4, Recharts, dnd-kit
Timeline
2025.09 — 2026.01
Live Project
—
The Challenge
The core problem
Leads scattered across email, phone, and SMS. Property data buried in separate systems. Seven-click workflows for basic actions. Agents were losing context and response speed—and generic CRMs were either overpriced or too rigid to handle real estate’s density of data.
Product vision
One workspace that combines lead management with deep property intelligence: ownership, tax, mortgage, price history, and comparables in one place.
What I owned
Designed scalable information architecture for high-density real estate data. Built 50+ reusable components and the full lead and property experience: filterable list with server-side pagination for 10k+ leads, tabbed property detail, comparison tools, notes and contacts. Reduced core workflows from 7 clicks to 3. Implemented Gmail OAuth and Twilio for email, SMS, and voice; added AI-assisted message generation. Shipped performance targets: sub-2s load, sub-500ms API, smooth at 10,000+ leads.
Results
Faster response time for agents. 60% less manual composition with AI-generated content. System holds up at 10k+ leads with sub-2s load and sub-500ms API. Showcased at LA Tech Week with 300 agents on the waiting list.
Design system
Atomic design with 50+ reusable components: color-coded status and source system, LeadCard, StatusBadge, ActionMenu, tabbed property panels (Features, Ownership, Price History, Tax, Mortgage, Lead Info), KPI cards, and 20+ workflow modals. Next.js 15 App Router, React 19, TypeScript, Tailwind 4; Recharts for price history and analytics; dnd-kit for sortable lists. API integration with service gateway and Bestia Property API for property modules and search.
Typography
Inter Bold
Inter Medium
Inter Regular
Colors
Primary
#1E40AF
Accent
#3B82F6
Highlight
#6366F1
Neutral
#FFFFFF

Leads list page
Central list view for all leads with status and source filters, full-text search, and server-side pagination. Each row shows key lead and property info with status/source badges; bulk actions and 3-click workflows for common tasks. Built to stay fast with 10,000+ leads. API routes support filtering by statusId, sourceId, search, page, and size.

Lead & property detail page
Single lead/property detail view with tabbed content: Property features, Ownership information, Price history (Recharts), Tax information, Mortgage, and Lead information. KPI section (Worth/Owe/Want), strategy notes, contacts, and property comparison with nearby search (Bestia Property API). Financial versions and notes/contacts backed by dedicated API routes. Clean, scannable layout for agents to move quickly from list to detail.
Email, SMS & Voice
Unified multi-channel communication: Gmail OAuth integration, Twilio for SMS and voice. Real-time call state, token refresh, and consistent UI for email/SMS composition. 60% reduction in manual composition with AI-generated content.

AI Assistants & Analytics
AI-powered content suggestions and automation; analytics dashboard for deal and team performance. 50+ reusable components, 106+ API routes; page load under 2s, API under 500ms. Launched at LA Tech Week with 300 agents on the waitlist.
Impact & results
Launched at LA Tech Week with 300 agents on the waiting list. Intuitive design, significant time savings, and positive early-adopter feedback. Real-time status and collaboration reduce overhead.
300
Agents on waitlist (LA Tech Week)
3
Clicks for common actions (was 7+)
60%
Less manual email/SMS time
<500ms
API response
Reflection
I’d push for more structured user interviews post–LA Tech Week to validate which workflows agents use most. On the technical side, I’d explore virtualized lists earlier for the 10k+ leads case. I learned how much clarity matters when you’re combining dense property data with multi-channel communication—one source of truth and a simple IA made the difference.